OBJECTIVE To study osseointegration of implants with surface modifications by the use of fluoroboric acid and/or H2 O2 installed in conventional sites or sites with circumferential marginal defects. MATERIAL AND METHODS Four implants with different surfaces were used. One basic surface (ZirTi(®) ) was sandblasted with zirconium microspheres and acid etched additionally with hydrofluoric acid. A second surface was treated with fluoroboric acid instead of hydrofluoric acid. The remainder of the other two surfaces was additionally treated with H2 O2 . The edentulous mandibles of 6 foxhound dogs were used to randomly install 8.5-mm-long implants with the different surfaces and to study the histological healing after 1 and 3 months. To study osteoconductivity, additional four recipient sites were prepared with the coronal region being widened so that a 4 mm deep and 0.85 mm wide marginal defect resulted after the placement of the four implants with different surfaces. No filler material or membranes were used, and a fully submerged healing was allowed for 3 months. RESULTS At the conventional sites, new bone formation ranged between 68.5% and 74.9% after 1 month. After 3 months, bone-to-implant contact ranged from 72.6% at the ZirTi(®) surface to 84.1% at the fluoroboric acid-treated implants, the difference being statistically significant. At the sites with marginal defects, bone formation ranged from 0.77 mm at the surface treated with fluoroboric acid and H2 O2 , to 1.93 mm at the surface treated with fluoroboric acid alone. CONCLUSIONS Fluoroboric acid treatment alone of titanium implant surfaces resulted in improved osseointegration and osteoconductivity after 3 months.